ALMA Observations of the Host Galaxy of GRB 090423 at Z = 8.23: Deep Limits on Obscured formation 630 Million Years after the .

The Harvard community has made this article openly available. Please share how this access benefits you. Your story matters

Citation Berger, E., B. A. Zauderer, R.-R. Chary, T. Laskar, R. Chornock, N. R. Tanvir, E. R. Stanway, A. J. Levan, E. M. Levesque, and J. E. Davies. 2014. "ALMA Observations of the Host Galaxy of GRB 090423 at Z = 8.23: Deep Limits on Obscured Star formation 630 Million Years after the Big Bang. ” The Astrophysical Journal 796 (2) (November 12): 96. doi:10.1088/0004-637x/796/2/96.

Published Version doi:10.1088/0004-637x/796/2/96

Citable link http://nrs.harvard.edu/urn-3:HUL.InstRepos:30498353

Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Open Access Policy Articles, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of- use#OAP arXiv:1408.2520v1 [astro-ph.GA] 11 Aug 2014 rde A018 USA 02138, MA bridge, cuee l 01 azlae l 01 n ta.2012; al. et Ono Ly 2011; narrow-band on al. or 2013), et 2011; Vanzella al. al. et et Finkelstein 2011; Bouwens 2009; al. al. usu- 2005; et al. et McLure (e.g., et Ouchi objects Malhotra 2006; 2004; confirmation al. al. bright et et Bouwens spectroscopic Bunker to 2003; subsequent al. limited et Stanway no therefore with are ally and LBGs), cpccnrain(.. oar ta.20;Iee l 200 al. spectro- et Iye subsequent 2003; al. requires et Kodaira also (e.g., confirmation which scopic LAEs), emitters; (ISM) medium a interstellar as galaxy well host properties as un- the , an of spectroscopic provide measurement of precise can spectroscopy and emission particular, ambiguous afterglow In of optical/near-IR epoch 2014). 2 the 2009, 2013, the al. al. et during Tanvir et 2006; Chornock including al. et Totani galaxy redshift, (e.g., and star high of at probes powerful formation as utilized be can (GRBs) ihro htmti esit teLy (the photometric al on et Laskar either (e.g., activity 2012). formation al. et star Tanvir 2011; th and as photometric such mass properties multi- study stellar to with galaxies host the combined insigh of unique follow-up be This 2013). then al. et can Chornock 2006; al. et Totani 20Es aionaBuead aaea A915 USA 91125, CA Pasadena, Boulevard, California East 1200 iyRa,LietrL17H UK 7RH, LE1 Leicester Road, sity oetyC47L UK 7AL, CV4 Coventry 2 3 4 5 1 bursts gamma-ray brightness, extreme their to Thanks ncnrs,drc aaysuisa hs esit rely redshifts these at studies galaxy direct contrast, In rpittpstuigL using typeset Preprint .B E. D pte cec etr aionaIsiueo Technolog of Institute California Center, Science Spitzer eateto hsc n srnm,Uiest fLeicest of University Astronomy, and Physics of Department eateto hsc,Uiest fWrik ibtHl R Hill Gibbet Warwick, of University Physics, of Department AA nvriyo ooaoUB39 ole,C 00,USA 80309, CO Boulder, 389, UCB Colorado of University CASA, avr-mtsna etrfrAtohsc,6 adnSt Garden 60 Astrophysics, for Center Harvard-Smithsonian ATVERSION RAFT LAO ALMA ERGER SFR otoeo edglxe at rate). galaxies formation field star of constant those and to Myr 100 of age population ecnld htGBhs aaisat constraining galaxies host most of GRB the host that the conclude is We than FIR UV/optical the rest-frame larger on limit our that ujc headings: Subject f for sample attractive formation. an are spectroscopy, afterglow from h ii nteuosue trfrainrt from rate formation star unobscured the on limit the 3 oprn omdlseta nrydsrbtosw lc a 22 place Arp we ULIRG distributions local energy the spectral of model luminosity to the Comparing than fainter times 20 otrmisudtce o3 to undetected remains host 8 . × 3fo h tcm ag ilmtrAry(LA n the and (ALMA) Array Millimeter Large Atacama the from 23 needn ftehs aaybrightness galaxy host the of independent epeetrs-rm a-nrrd(I)adotclobser optical and (FIR) far- rest-frame present We 10 1 UV .A Z A. B. , 10 SRAIN FTHE OF BSERVATIONS . A L A UGUST ⊙ T M 1 E tl mltajv 03/07/07 v. emulateapj style X orsodn oalmto h bcrdsa omto rat formation star obscured the on limit a to corresponding , 1. AUDERER ⊙ INTRODUCTION 3 2014 13, yr − GB090423) (GRB aais ihrdhf ai otnu:glxe —gamm galaxies continuum: radio — high-redshift galaxies: 1 eas lc ii ntehs aayselrms of mass stellar galaxy host the on limit a place also We . 1 .R C R.-R. , F ORMATION z σ & H iisof limits OST Lmnbekglxe,Ly galaxies, break (Lyman 4 α HARY ra technique; break G α .M L M. E. 1,2 LX OF ALAXY 3 M 630 mgn (Ly imaging F z .L T. , ν ,M 220-6, MC y, & (222GHz) rf eso uut1,2014 13, August version Draft r Univer- er, ,epcal hs ihmaue neselrmdu meta medium interstellar measured with those especially 4, et Cam- reet, EVESQUE ILLION ASKAR (e.g., 012; ABSTRACT oad, eir R 090423 GRB 6; α t . 1 . .C R. , Y 5 ubeSaeTelescope Space Hubble .E D E. J. & , 33 EARS aeo As h eineo Ly the on In reliance 2012). al. the et Shibuya LAEs, 2010; of al. et case Ouchi 2010; al. et Hu oiyof nosity iiylvlo nyafwM sens few a probe a to to only galaxies time of high-redshift level first tivity in the formation for star possible obscured it makes (ALMA) the Array of bulk sample the attractive an observations. on provide follow-up multi-wavelength redshift high GRB at shed spectroscopically-confirmed galaxies ex- host not Thus, are do population. galaxies galaxy and forming ex- star rare these highly However, ceedingly and 2013). al. luminous et tremely Riechers [C 2012; and al. et CO Walter via hig few confirmation spec a with scopic of (SMGs), identification submillimeter galaxies the of submillimeter-bright to redshift led technique also alternative has redshif since selection An are color instrumentation mid-IR. lines proper- current the emission ISM of into optical the rest-frame reach relevant of the the beyond measurements for are Moreover, LAEs, ties and environments. LBGs dusty both against selection Ly LAE and UV rest-frame on ii eetosdrn h pc frinzto u osig to due Ly of reionization absorption of and epoch scattering nificant the during detections hibit ue S ealcte ilse ih nterltosbe- formati relations star the obscured and on content, light long mea- dust shed the , with will tween galaxies In host ISM GRB 2013). sured of al. et observations Riechers ALMA 2012; term, al. et Walter 2012; L yr n orsodnl ihsa omto ae of rates formation star high correspondingly and iso eeto fteLy the re- of also th detection which inhibits confirmation, on also spectroscopic lies effect LBG same of the success medium; intergalactic tral R 943b iteo hi eeto techniques. selection their of virtue by 090423 GRB µ HORNOCK tr LAsuiso ihrdhf bcrdstar obscured redshift high of studies ALMA uture IR yand Jy − nti otx,teavn fteAaaaLreMillimeter Large Atacama the of advent the context, this In α 1 ain ftehs aayo R 943at 090423 GRB of galaxy host the of vations ii nteI uioiyof luminosity IR the on limit ial,w opr u ilmtrobservations millimeter our compare we Finally, A & odt,atog h edglxe aemuch have galaxies field the although date, to mtes n umliee aais,adfind and galaxies), submillimeter and emitters, eeacsil eg,Cxe l 01 obse al. et Combes 2011; al. et Cox (e.g., accessible were TRTHE FTER ,adcmaal otelclsabrtM82. M starburst local the to comparable and 0, 10 AT AVIES F pte pc Space Spitzer 12 ν L z 1 (3 IR 8 = .R T R. N. , fSFR of e L 1 . ⊙ 6 ∼ . µ 3 D 23: utalmnu nrrdglxe;ULIRGs) galaxies; infrared (ultra-luminous B 10 m) IG 10 ANVIR . IR M B EEP etfaeU bevtosis observations UV rest-frame L 1ny h I ii sabout is limit FIR The nJy. 81 . ∗ ANG ⊙ . rvosy nyrr Mswith SMGs rare only Previously, . M 5 α L 3 5 .R S R. E. , MT ON IMITS ⊙ msinba ohteLGand LBG the both bias emission -a us:individual burst: a-ray α × ⊙ yr 10 yr ie iial,tereliance the Similarly, line. − L epciey The respectively. , − 1 7 IR 1 ra nertdI lumi- IR integrated an or , M o comparison, for ; TANWAY α (8 II ⊙ msinmyas in- also may emission O − msinlns(e.g., lines emission ] α frastellar a (for BSCURED 1000 htn yteneu- the by photons 4 .J L J. A. , llicities µ m) z . ∼ = S EVAN TAR 10 3 M 4 tro- ted for , on ⊙ i- h e - 2 at redshifts that are inaccessible with other techniques (i.e., source location of −4 ± 11 µJy), leading an upper limit of z & 4). This will also provide a comparison with the rest- Fν(222GHz) . 33 µJy (3σ; Table 1). frame FIR properties of LBGs and LAEs, although none have been individually detected to date at z & 4 (e.g., Ouchi et al. 2.2. Spitzer Space Telescope 2013; Ota et al. 2014). Spitzer observations were obtained during the warm mis- Here, we present deep ALMA observations of the host sion (DDT Program 538) with the Infrared Array Camera galaxy of the most distant spectroscopically-confirmed burst (IRAC) at 3.6 µm on 2010 January 26 and 27 UT. Each obser- to date – GRB090423 at z = 8.23 (Salvaterra et al. 2009; vation consisted of 650 100-s frames (93.6 s on source), for a Tanvir et al. 2009). The observations are aimed at reaching combined on-source time of 33.8 hr. We analyzed the data us- a rest-frame FIR luminosity level below that of a luminous in- ing the Great Observatories Origins Deep Survey (GOODS) 11 frared galaxy (LIRG: LIR = 10 L⊙). We also present Spitzer pipeline, including corrections for bright star image artifacts Space Telescope rest-frame optical observations to constrain (muxbleed and pulldown) with dark and sky background sub- the stellar mass of the host galaxy. The ALMA and Spitzer traction. We carried out astrometric alignment utilizing observations and data analysis are described in §2. In §3 we in common with the Sloan Digital Survey Survey database, use the data to constrain the host galaxy properties, while in and created a combined mosaic of the individual frames on a §4 we compare these results to millimeter studies of LAEs, 0.4′′ × 0.4′′ pixel grid. We also produced a pair of split im- LBGs, and SMGs at z & 4. ages, each with half of the frames, and used the difference in these split images to produce a noise map. The resulting un- 2. OBSERVATIONS certainty is about 2 times higher than that derived from pure The discovery and redshift determination of GRB090423 pixel statistics in the vicinity of the source. +0.06 are presented in Tanvir et al. (2009), who find z =8.23−0.07. We do not detect significant emission at the location of The position of the afterglow is R.A.=09h55m33.29s, GRB090423 (Figure 1). There is no evidence for signifi- Decl.= +18◦08′57.8′′ (J2000) with an uncertainty of about cant source confusion at this location, in agreement with the 0′′.1 in each coordinate. The host galaxy was previously HST/WFC3 images of the field (Tanvir et al. 2012). From the observed with the Hubble Space Telescope Wide-field Cam- noise maps, we find that the 1σ point source sensitivity at the era 3 (HST/WFC3) in the F110W and F160W filters, leading location of GRB090423 is 27 nJy. in a 1.2” ra- to non-detections (Tanvir et al. 2012). It was also observed dius aperture yields a 3.6 µm flux density of 26±27 nJy, after with the Australia Telescope Compact Array (ATCA), yield- aperture corrections. We therefore place an upper limit on the ing non-detections of rest-frame 850 µm continuum emission brightness of the host of Fν (3.6µm . 81nJy (3σ; Table 1). and CO (3-2) emission (Stanway et al. 2011). The continuum limits are summarized in Table 1. 3. HOST GALAXY PROPERTIES: IR LUMINOSITY, OBSCURED STAR FORMATION RATE, AND STELLAR MASS 2.1. Atacama Large Millimeter Array The host galaxy of GRB090423 remains undetected in our ALMA band 6 (222 GHz) observations were carried out ALMA and Spitzer observations. The flux density limits with 25 − 28 12-m diameter antennas in five observing blocks are plotted in the left panel of Figure 2. Converted to rest- spread over 12 days on 2013 November 18, 21, and 30 UT frame spectral luminosity, the non-detections correspond to × 31 −1 −1 (Cycle 1). The 2 GHz wide spectral windows were set to cen- Lν (145µm) . 3.1 10 erg s Hz and Lν (0.39µm) . × 28 −1 −1 tral frequencies of 213, 215, 229 and 231 GHz. The total 7.6 10 erg s Hz , respectively (right panel of Figure 2, on-source integration time was 163 min after data flagging. and Table 1). We performed data calibration and imaging using the Com- Also shown in Figure 2 are the spectral energy distribu- mon Astronomy Software Application (CASA). The gain tions (SEDs) of several local galaxies redshifted to z =8.23: calibration utilized J0854+2006, while bandpass calibration the ULIRG Arp 220, the starburst M 82, the low metallicity utilized one of J0522−3627, J0538−4405, J1037−2934, or dwarf galaxy IZw18, and the host galaxy of GRB980425 J1058+0133 in each observing block. Absolute flux calibra- (d = 40 Mpc; Michałowskiet al. 2014). Our ALMA non- tion was performed using observations of the solar system ob- detection corresponds to a luminosity that is about 20 times jects Ceres, Ganymede, or Pallas. We found some minor vari- fainter than that of Arp220 and about 1.6 times higher than ations in the resulting fluxes of J0854+2006 between the ob- that of M82. We also provide a comparison with the galaxy serving blocks, likely due to the use of asteroids, which have templatesof Rieke et al. (2009), which are based on eleven lo- uncertain flux models. We therefore used Ganymede when cal star forming LIRGs and ULIRGs, and find that the ALMA × 10 possible and Ceres to determine and manually set the flux for non-detection corresponds to LIR . 3 10 L⊙. J0854+2006 in each epoch6. The overall uncertainty in the For the Arp220 template, the ALMA non-detection pro- absolute flux calibration is about 10%. vides a much more stringent limit than the HST and Spitzer The resulting combined continuum map (natural weight- limits, which only rule out a luminosity comparable to that of ing without primary beam correction) has an rms noise of 11 Arp220. The ALMA constraints are also more stringent for µJy beam−1 for a synthesized beam size (full-width at half the M82 template, although only by about a factor of 2. For maximum) of 1′′.0 × 0′′.80 (Figure 1). No source is detected an Sd galaxy template from the Spitzer Wide-area InfraRed at the location of GRB0904237 (with a flux density at the Extragalactic survey (SWIRE) library, the ALMA, HST, and Spitzer data all place comparable limits on the host SED. On 6 We used the following values for each spectral window and date: the other hand, for the IZw18 template, which has weak FIR Nov. 18: 2.43, 2.39, 2.29, and 2.28 Jy; Nov. 21: 2.25, 2.24, 2.17, and 2.16 Jy; emission, the HST limits place a much more stringent con- and Nov. 30: 1.81, 1.80, 1.74, and 1.74 Jy. straint than the ALMA data, by about a factor of 600. The 7 h m s We note the presence of a 3σ peak at R.A. =09 55 33.17 , same is true for the GRB980425 host galaxy template, which Decl. =+18◦08′59.0′′ (J2000), about 2′′ away from the position of GRB 090423, but given the low signal-to-noise ratio and the substantial offset is about 15 times fainter than the ALMA limit when scaled to we do not consider this potential source to be related to GRB 090423. the HST limits. 3

Using the standard modified blackbody SED of dust emis- (Figure 3). sion, with a dust temperature range of Tdust ≈ 30 − 50 K and Expanding the redshift range down to z = 4, only a few β ≈ 1.5 − 2 (e.g., Swinbanket al. 2014), the ALMA non- galaxies with spectroscopic redshifts have been detected detection corresponds to an upper limit on the integrated IR with ALMA, all of which are bright SMGs (Figure 3). 10 luminosity of LIR(8−1000µm) . (2−5)×10 L⊙; the lower Swinbanket al. (2012) detected two SMGs in ALMA ob- bound on the blackbody temperature is set by the cosmic servations of the LABOCA Extended Chandra Deep Field- microwave background temperature TCMB(z =8.23) ≈ 25 K South Survey, with [C II]λ157.74 µm detections indicating (da Cunha et al. 2013). The IR luminosity upper limit is thus redshifts of z =4.419 and z =4.444, and resulting IR lumi- 12 a few times lower than the scale for a LIRG, and agrees well nosities of 2 × 10 L⊙. Wagg et al. (2012) and Carilli et al. with the comparison to the Rieke et al. (2009) templates. The (2013) presented continuum and [C II] line emission from the limit on the integrated IR luminosity corresponds to an upper SMG and two LAEs associated with the AGN/SMG system bound on the obscured star formation rate (Kennicutt 1998; BRI1202-0725 at z =4.70, with luminosities of 1.2 × 1013 8 −1 11 12 Calzetti et al. 2010), of SFRIR . 3 − 5M⊙ yr . For compar- L⊙ (SMG), & 3.6 × 10 L⊙ (Lyα-1), and 1.7 × 10 L⊙ ison, the limit on the unobscured star formation rate from the (Lyα-2). Hezaveh et al. (2013) detected two strongly lensed −1 HST rest-frame UV non-detection is SFRUV . 1.2M⊙ yr . SMGs at z =4.224 and z =5.656, found in South Pole Tele- We also place an upper bound on the host galaxy stellar scope data, with intrinsic IR luminosities of about 3.8 × 1012 13 mass using the HST and Spitzer upper limits. Utilizing the and 3.7 × 10 L⊙, respectively. Wiklind et al. (2014) de- Bruzual & Charlot (2003) synthesis models tected an SMG at z =4.762, first identified in LABOCA ob- with a constant star formation rate, a Salpeter IMF, a metallic- servations of the Extended Chandra Deep Field, with LIR ≈ 13 ity of 0.2Z⊙, and no dust extinction, we find that for a stellar 1.3 × 10 . Riechers et al. (2014) detected the SMG AzTEC- 13 population age of 10−100 Myr (typical of GRB host galaxies; 3 at z =5.299 with LIR ≈ 1.7 × 10 L⊙, and placed a limit on Savaglio et al. 2009; Leibler & Berger 2010) the limit on the an LBG in the same field at a similar redshift of z =5.295 of 7 11 host stellar mass is M∗ . (1 − 5) × 10 M⊙. The inferred host LIR . 5.3 × 10 L⊙. galaxy properties are summarized in Table 2. In addition to the ALMA observations, Cox et al. (2011) studied ID141, a lensed galaxy at z =4.243 from the Her- 4. COMPARISON TO MILLIMETER OBSERVATIONS OF FIELD schel Astrophysical Terahertz Large Area Survey, and in- 12 GALAXIES AT Z & 4 ferred LIR ≈ (3 − 8) × 10 L⊙; the range of values accounts To place the non-detection of the host galaxy of for the unknown lensing magnification factor. Combes et al. GRB090423 in the context of millimeter studies of high- (2012) discovered and studied a lensed SMG at z = 5.243 ≈ 13 redshift galaxies, we summarize below previous observational from the Herschel Lensing Survey with LIR 10 L⊙. efforts to detect and study distant LAEs, LBGs, and SMGs; Walter et al. (2012) determined a redshift of z = 5.183 for the SMG HDF850.1 based on CO and [C II] line emission, the information is summarized in Figure 3. We first note 12 leading to LIR ≈ 8.7 × 10 L⊙. Davies etal. (2012) used that, at present, ALMA observations of only two other galax- ∼ ies with spectroscopic redshifts of z & 6 have been pub- stacked MAMBO-2 observations of z 5 LBGs to place a × 11 lished: the LAEs “Himiko” at z =6.595 (Ouchi et al. 2013; limit on their mean luminosity of LIR . 3 10 L⊙, while Ota et al. 2014) and IOK-1 at z =6.96 (Ota et al. 2014). Nei- Coppin et al. (2014) used stacked SCUBA-2 observations of ∼ ∼ 12 ther source was detected, with resulting 3σ flux density lim- z 4.8 LBGs to find a mean luminosity of LIR 10 L⊙ its of Fν(259GHz) . 52 and Fν (232GHz) . 63 µJy, respec- (based on a 3.8σ detection at 850 µm). tively. The flux density and integrated IR luminosity lim- Finally, two other GRB host galaxies at lower redshifts have its for these two LAEs are shallower than those achieved in been observed with ALMA (Wang et al. 2012). The host of our observations; see Figure 3. However, it is important to GRB021004 at z =2.330 was not detected, with a resulting × 11 −1 note that Himiko and IOK-1 are significantly more luminous limit of LIR . 3 10 L⊙ (SFRIR . 50M⊙ yr ). Thehostof the dusty GRB080607 at z =3.036 was marginally detected than the host of GRB090423 in the rest-frame UV and opti- 11 (3.4σ) with a resulting luminosity of (2.4 − 4.5) × 10 L⊙ cal, by a factor of & 30 (Figure 2), with inferred unobscured −1 −1 (SFR ≈ 40 − 80 M⊙ yr ). Prior to ALMA, a few GRB host star formation rates of ≈ 25 − 30 M⊙ yr (Ouchi et al. 2013; IR Otaetal. 2014). Thus, in terms of the ratio of FIR to UV galaxies at z . 2 were detected in the submillimeter and radio luminosity, the limits for Himiko and IOK-1 are more con- bands(Berger et al.2001; Frail et al.2002; Berger et al. 2003; straining than those for the host of GRB090423. Michałowski et al. 2014; Symeonidis et al. 2014), some with Beyond the ALMA observations of z & 6 galaxies, there inferred IR luminosities that exceed that of Arp220. How- are shallower observations from other millimeter facilities. ever, most GRB host galaxies were not detected at a compara- Booneet al. (2007) observed the lensed LAE HCM6A at ble level to Arp220 or fainter (Berger et al. 2003; Tanvir et al. z =6.56 with the Plateau de Bure Interferometer and placed 2004; Priddey et al. 2006; Perley & Perley 2013), ruling out a 11 dominant ULIRG host population. a limit of LIR . 2.1 × 10 L⊙. González-López et al. (2014) used the Combined Array for Research in Millimeter-wave Thus, in comparison to the previous studies of z & 4 Astronomy to place limits on two LAEs at z = 6.541 and spectroscopically-confirmed field galaxies, as well as other 12 12 GRB host galaxies, our ALMA observations of the host of z = 6.554 of . 1.0 × 10 and 2.1 × 10 L⊙, respectively. Riechers et al. (2013) used a wide range of facilities to study GRB090423 represent the deepest limit to date in terms of the bright SMG HFLS3 at z =6.337, discovered in the Her- flux density, spectral luminosity, and integrated IR luminosity (Figure 3). schel Multi-tiered Extragalactic Survey, and found LIR ≈ × 13 2.9 10 L⊙. Thus, the only galaxy at z & 6 detected in the 5. CONCLUSIONS millimeter band to date is a hyper-luminous infrared galaxy We present ALMA and Spitzer observations of the host 8 Using the scaled SED of I Zw 18, which has much weaker FIR emission galaxy of GRB090423 at z = 8.23, the highest redshift −1 relative to its optical/UV emission, leads to a limit of SFRIR . 100 M⊙ yr . spectroscopically-confirmed galaxy observed with these fa- 4 cilities to date. The host galaxy remains undetected at rest- and with a subset of the ALMA antennas. On the other hand, frame wavelength of 145 µm (ALMA) and 0.39 µm(Spitzer). if typical high redshift galaxies have SEDs similar to the lo- The resulting limit on the integrated IR luminosity is LIR . cal dwarf IZw18 or the host galaxy of GRB980425, then 10 −1 3 × 10 L⊙, corresponding to . 5 M⊙ yr of obscured much deeper ALMA observations will be required to detect star formation; scaling the SED of the low metallicity dwarf their FIR emission. Still, looking forward we anticipate that galaxy IZw18 relaxes the star formation rate upper limit to ALMA observations of GRB host galaxies will be highly de- −1 . 100 M⊙ yr . In addition, based on the Spitzer and HST sirable and productive, especially targeting spectroscopically- non-detectionswe place a limit on the host galaxy stellar mass confirmed hosts at z & 4 with detailed ISM metallicity mea- 7 of M∗ . 5×10 M⊙ (100 Myr old stellar population with con- surements, a unique sample among high-redshift galaxies. stant star formation rate). The limit on the unobscured star formation rate based on HST rest-frame UV observations is −1 . 1.2M⊙ yr . The Berger GRB group at Harvard is supported in part We additionally compare our ALMA non-detection to mil- by the National Science Foundation under Grant AST- limeter observations of spectroscopically-confirmed galaxies 1107973. This paper makes use of the following ALMA at z & 4 (undertaken with ALMA and other facilities) and data: ADS/JAO.ALMA#2012.1.00953.S.ALMA is a partner- show that the limit on the host of GRB090423 is the deep- ship of ESO (representing its member states), NSF (USA) est to date compared to any published observations. At this and NINS (Japan), together with NRC (Canada) and NSC redshift range only SMGs have been convincingly detected, and ASIAA (Taiwan), in cooperation with the Republic of while individual LBGs and LAEs have so far escaped de- Chile. The Joint ALMA Observatory is operated by ESO, tection even with ALMA (a marginal detection of LBGs at AUI/NRAO and NAOJ. The National Radio Astronomy Ob- z ∼ 4.8 in stacked SCUBA-2 observations has been reported servatory is a facility of the National Science Foundation op- by Coppin et al. 2014). The only comparable limits to ours erated under cooperative agreement by Associated Universi- are based on ALMA observations of two LAEs at z =6.595 ties, Inc. This work is based in part on observations made and z =6.96. with the Spitzer Space Telescope, which is operated by the It is quite remarkable that we were able to reach a bright- Jet Propulsion Laboratory, California Institute of Technology ness limit of about 20 times lower than Arp220 (and compa- under a contract with NASA. rable to M82) at z =8.23 in only 3 hours of on-source time Facilities: ALMA, Spitzer (IRAC)

REFERENCES Berger, E., Cowie, L. L., Kulkarni, S. R., Frail, D. A., Aussel, H., & Barger, McLure, R. J., et al. 2011, MNRAS, 418, 2074 A. J. 2003, ApJ, 588, 99 Michałowski, M. J., et al. 2014, A&A, 562, A70 Berger, E., Kulkarni, S. R., & Frail, D. A. 2001, ApJ, 560, 652 Ono, Y., et al. 2012, ApJ, 744, 83 Boone, F., Schaerer, D., Pelló, R., Combes, F., & Egami, E. 2007, A&A, 475, Ota, K., et al. 2014, ArXiv e-prints 513 Ouchi, M., et al. 2013, ApJ, 778, 102 Bouwens, R. J., Illingworth, G. D., Blakeslee, J. P., & Franx, M. 2006, ApJ, Ouchi, M., et al. 2009, ApJ, 706, 1136 653, 53 Ouchi, M., et al. 2010, ApJ, 723, 869 Bouwens, R. J., et al. 2011, ApJ, 737, 90 Perley, D. A., & Perley, R. A. 2013, ApJ, 778, 172 Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000 Priddey, R. S., Tanvir, N. R., Levan, A. J., Fruchter, A. S., Kouveliotou, C., Bunker, A. J., Stanway, E. R., Ellis, R. S., & McMahon, R. G. 2004, MNRAS, Smith, I. A., & Wijers, R. A. M. J. 2006, MNRAS, 369, 1189 355, 374 Riechers, D. A., et al. 2013, Nature, 496, 329 Calzetti, D., et al. 2010, ApJ, 714, 1256 Riechers, D. A., et al. 2014, ArXiv e-prints Carilli, C. L., Riechers, D., Walter, F., Maiolino, R., Wagg, J., Lentati, L., Rieke, G. H., Alonso-Herrero, A., Weiner, B. J., Pérez-González, P. G., McMahon, R., & Wolfe, A. 2013, ApJ, 763, 120 Blaylock, M., Donley, J. L., & Marcillac, D. 2009, ApJ, 692, 556 Chornock, R., Berger, E., Fox, D. B., Fong, W., Laskar, T., & Roth, K. C. Salvaterra, R., et al. 2009, Nature, 461, 1258 2014, ArXiv e-prints Savaglio, S., Glazebrook, K., & Le Borgne, D. 2009, ApJ, 691, 182 Chornock, R., Berger, E., Fox, D. B., Lunnan, R., Drout, M. R., Fong, W.-f., Shibuya, T., Kashikawa, N., Ota, K., Iye, M., Ouchi, M., Furusawa, H., Laskar, T., & Roth, K. C. 2013, ApJ, 774, 26 Shimasaku, K., & Hattori, T. 2012, ApJ, 752, 114 Combes, F., et al. 2012, A&A, 538, L4 Stanway, E. R., Bremer, M. N., Tanvir, N. R., Levan, A. J., & Davies, L. J. M. Coppin, K. E. K., et al. 2014, ArXiv e-prints 2011, MNRAS, 410, 1496 Cox, P., et al. 2011, ApJ, 740, 63 Stanway, E. R., Bunker, A. J., & McMahon, R. G. 2003, MNRAS, 342, 439 da Cunha, E., et al. 2013, ApJ, 766, 13 Swinbank, A. M., et al. 2012, MNRAS, 427, 1066 Davies, L. J. M., Bremer, M. N., Stanway, E. R., Mannering, E., Lehnert, Swinbank, A. M., et al. 2014, MNRAS, 438, 1267 M. D., & Omont, A. 2012, MNRAS, 425, 153 Symeonidis, M., et al. 2014, ArXiv e-prints Finkelstein, S. L., et al. 2013, Nature, 502, 524 Tanvir, N. R., et al. 2004, MNRAS, 352, 1073 Frail, D. A., et al. 2002, ApJ, 565, 829 Tanvir, N. R., et al. 2009, Nature, 461, 1254 González-López, J., et al. 2014, ApJ, 784, 99 Tanvir, N. R., et al. 2012, ApJ, 754, 46 Hezaveh, Y. D., et al. 2013, ApJ, 767, 132 Totani, T., Kawai, N., Kosugi, G., Aoki, K., Yamada, T., Iye, M., Ohta, K., & Hu, E. M., Cowie, L. L., Barger, A. J., Capak, P., Kakazu, Y., & Trouille, L. Hattori, T. 2006, PASJ, 58, 485 2010, ApJ, 725, 394 Vanzella, E., et al. 2011, ApJ, 730, L35 Iye, M., et al. 2006, Nature, 443, 186 Wagg, J., et al. 2012, ApJ, 752, L30 Kennicutt, R. C., Jr. 1998, ARA&A, 36, 189 Walter, F., et al. 2012, Nature, 486, 233 Kodaira, K., et al. 2003, PASJ, 55, L17 Wang, W.-H., Chen, H.-W., & Huang, K.-Y. 2012, ApJ, 761, L32 Laskar, T., Berger, E., & Chary, R.-R. 2011, ApJ, 739, 1 Wiklind, T., et al. 2014, ApJ, 785, 111 Leibler, C. N., & Berger, E. 2010, ApJ, 725, 1202 Malhotra, S., et al. 2005, ApJ, 626, 666 5

TABLE 1 OBSERVATIONS OF THE HOST GALAXYOF GRB090423

a a Instrument λobs Fν Lν Reference (µJy) (erg s−1 Hz−1) ATCA 0.79 cm . 9.3 . 8.7 × 1030 Stanway et al. 2011 ALMA 0.14 cm . 33 . 3.1 × 1031 This paper Spitzer/IRAC 3.6 µm . 8.1 × 10−2 . 7.6 × 1028 This paper HST/WFC3 1.54 µm . 9.1 × 10−3 . 8.6 × 1027 Tanvir et al. 2012 HST/WFC3 1.25 µm . 6.0 × 10−3 . 5.6 × 1027 Tanvir et al. 2012

NOTE. — a Limits are 3σ. 6

TABLE 2 INFERRED PROPERTIES OF THE HOST GALAXYOF GRB090423

Parameter Value

a 10 LIR . (2 − 5) × 10 L⊙ a −1 SFRIR . 3 − 5 M⊙ yr b −1 SFRUV . 1.2 M⊙ yr c 7 M∗ . (1 − 5) × 10 M⊙ a NOTE. — Assuming Tdust ≈ 30 − 50 K and β ≈ 1.5 − 2. b host Assuming AV = 0 mag. c Assuming a stellar population age of 10−100 Myr, constant star forma- tion rate, Salpeter IMF, Z = 0.2Z⊙, host and AV = 0 mag. 7

ALMA Spitzer

N

E

FIG. 1.— Left: ALMA band 6 continuum image of a 20′′ × 20′′ region centered on the location of the host galaxy of GRB 090423 (cross). Contours are in steps of 1σ = 11 µJy beam−1 starting at ±2σ (solid: positive; dashed: negative). No millimeter emission is detected at the location of the host galaxy. Right: Spitzer/IRAC 3.6 µm image of a 20′′ × 20′′ region centered on the location of the host galaxy of GRB 090423 (cross). No infrared emission is detected at the location of the host galaxy. 8

2 32 10 10 ALMA GRB 080607 (z=3.036) GRB 021004 (z=2.330) 1 ATCA 31 10 10 Himiko (z=6.595) IOK−1 (z=6.96)

0 30 10 10 Hz) Spitzer/IRAC /

−1 s 29 Jy)

10 / 10 µ ( HST/WFC3 ν −2 (erg 28 F 10 10 ν L

−3 27 10 10 Arp 220 (/20) M 82 (x1.6) Arp 220 (/20) M 82 (x1.6) −4 I Zw 18 (x10) GRB 980425 (x3) 26 I Zw 18 (x10) GRB 980425 (x3) 10 Sd template (SWIRE library) 10 Sd template (SWIRE library) 10 L(FIR) = 3x10 L (Rieke+2009) L(FIR) = 3x1010 L (Rieke+2009) o o 0 1 2 3 4 −1 0 1 2 3 10 10 10 10 10 10 10 10 10 10 λob s (µm) λr es t (µm)

FIG. 2.— Left: Limits on the flux density of the host galaxy of GRB 090423 in the near-IR (HST), mid-IR (Spitzer), millimeter (ALMA), and radio (ATCA). Also shown are the SEDs of the local ULIRG Arp 220 (red), the local starburst M 82 (blue), the local dwarf I Zw 18 (green), the local host galaxy of GRB 980425 10 (gray; Michałowski et al. 2014), an Sd galaxy template (magenta), and a template for a galaxy with LIR = 3 × 10 L⊙ (cyan; Rieke et al. 2009) all shifted to z = 8.23; with the exception of I Zw 18 and the host of GRB 980425, which are scaled to the HST limits, the galaxy models are scaled to match the ALMA flux density limit. For the Arp 220 template the ALMA non-detection places a stronger constraint on the SED than the HST and Spitzer limits, while for the starburst templates the limits are comparable. For the I Zw 18 and GRB 980425 host galaxy templates, the HST limits are more constraining. Right: Same as the left panel, but plotting the rest-frame luminosity density and wavelength. Also shown are the ALMA observations and rest-frame UV/optical SEDs of two other GRB host galaxies (circles: detections; triangles: upper limits; GRB 080607 is a marginal 3.4σ detection; Wang et al. 2012), and two spectroscopically-confirmed LAEs at z ≈ 6.6 − 7.0 (Ouchi et al. 2013; Ota et al. 2014). 9

4 10 SPT−S J034640−5205.1 HFLS3

AzTEC−3 13 LESS73 HLSJ091828.6+514223 10 BRI1202−0725 (SMG) HDF850.1 3 ID141 10 SPT−S J041840−4752.0

ALESS61.1 yr) SDF J132408.3 ) BRI1202−0725 (LAE−2) / ⊙ ALESS65.1 ⊙ 12 LBG stack SDF J132415.7

(L 10 LBG stack 2 (M LBG−1 10 IR

BRI1202−0725 (LAE−1) IR L

HCM 6A

11 SFR 10 GRB host Himiko 1 LAEs IOK−1 10 GRB 090423 LBGs SMGs 10 10 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 Redshift

−44 −1 FIG. 3.— Integrated IR luminosities and inferred obscured star formation rates (SFRIR = 4.5 × 10 LIR M⊙ yr ; Kennicutt 1998) of galaxies with spectro- scopic redshifts of z & 4 observed with ALMA and other millimeter (circles: detections; triangles: upper limits). The objects include SMGs (cyan), LBGs (green), LAEs (red), and the host of GRB 090423 (black). So far, mostly SMGs have been detected as individual galaxies at z & 4. See §4 for details and references.